Not sure if this is right or not....
To make the "down" bias setting sound OK I have changed out the 120K resistor with 10K. I am talking about the 180K/120K voltage divider that feeds the inverter.
This seems a bit drastic to me, but seems to work. Did anyone else have to make this radical a value change? I have gone over the circuit and can't find anything else wrong. Note that I am not using batteries, so I can't see how this value change is going to "hurt" anything.
I am using 47uF instead of 4.7uF for the decay cap and 4.7K for Rx. Other than that stock values are used everywhere. I am OK with the sweep but not the decay time, so I may mess with this some more. but I can't see how these values will affect the down bias in any event.
I am using the special 1 LED + 2 resistor can from small bear for the opto.
Update. I spent some more time on this pedal yesterday.
--I am finding an interaction between the "down" bias and the value for Rx. Since the Rx value affects half the inverting side of op amp U3a I guess this makes sense.
--I have put in a 22uF cap for c9, a 10K multiturn trimmer for Rx (dangerous--don't want to make the - of the opto see ground or maybe it'll blow up) and a 100k trimmer for R19. So far this has yielded the most musical results. I am surprised how small the "sweet spot" is for the best match between up and down bias. However, I am still not 100% happy with the sweep, esp for the down direction.
I am next going to work on knocking down some of the gain--the gain stage gives me a lot more gain then I need right now. After that hopefully I'll be done....
BTW in case anyone out there is still listening....I have another question--why is C1 spec'd as tantulum? Isn't using tant caps in the signal chain a no-no because they sound crappy? Can a electro be used here instead w/o affecting the quality of the sound?